To identify cellular interacting partners of SARS-CoV accessory protein 3b, yeast two-hybrid screening of human lung cDNA library with full-length 3b as bait was conducted. The bait plasmid was constructed by cloning 3b coding sequence in-frame with the lex A DNA binding domain (Fig. 1A). The human lung cDNA library, cloned in-frame with the B42 activation domain vector was purchased commercially. From the screening, a clone encoding 51–421 aa of Runt related transcription factor 1, isoform b (RUNX1b, accession no. NP_001001890) (Fig. 1A) was obtained. Yeast co-transformants containing pYesTrp2-RUNX1b (51–421 aa) and pHybLexA/Zeo-3b withstood stringent growth conditions of media (supplemented with 5 mM 3-Aminotrizol) and scored positive for β–galactosidase filter lift assay (Fig. 1B) whereas co-transformants of pHybLexA/Zeo and pYesTrp2 (empty bait and prey plasmids); pHybLexA/Zeo-3b and pYesTrp2 (bait plasmid with empty prey plasmid); and pYesTrp2-RUNX1b and pHybLexA/Zeo (prey plasmid with empty bait plasmid) did not. pYesTrp2-Fos and pHybLexA/Zeo-Jun co-transformed yeast colonies were used as positive control.

RUNX1b and 3b physical interaction was confirmed in vitro by co-immunoprecipitation in two separate experiments. Firstly, co-immunoprecipitation using in vitro transcribed and translated full-length [³⁵S]-3b and [³⁵S]-RUNX1b showed pull-down of 3b by anti-RUNX1 from 3b-RUNX1b lysate mixture whereas no corresponding band for 3b was visible from the 3b and RUNX1b lysates alone (Fig. 1C). The interaction was further confirmed in Huh7 cells. Cells expressing flag tagged RUNX1b (pCMV-Tag2B Flag RUNX1b or Flag-RUNX1b) and myc tagged 3b (pCDNA 3.1 Myc/His-3b or Myc-3b), alone or together, were subjected to co-immunoprecipitation assay. Immunoprecipitation by anti-Flag detected 3b from cells co-expressing RUNX1b and 3b (Lane 3, Fig. 1D) whereas no corresponding band for 3b was seen from cells expressing RUNX1b or 3b alone (Lane 1 and 2, Fig. 1D). Reciprocal co-immunoprecipitation using anti-GFP antibody in cells expressing EGFP, 3b-EGFP, 9b-EGFP (negative control) and RUNX1b expression constructs showed pull-down of RUNX1b with 3b specifically, confirmed physical interaction of 3b with RUNX1b.

Cellular distribution of full-length 3b and RUNX1b was visualized in HEK293 cells using immunofluorescence assay. HEK293 cells transfected with Flag-RUNX1b and HA-3b (pXJ40-HA-3b) expression plasmids were subjected to immunofluorescence assay using anti-HA and anti-RUNX1 antibodies. Confocal microscopy revealed localization of RUNX1b in the nucleus, as reported earlier [19] and 3b in the nucleus with majority inside the nucleolus, as seen by Yuan et al [11]. Co-transfected cells expressing 3b and RUNX1b showed significant partial co-localization of the two proteins in the extra-nucleolar nucleus area with Pearson's correlation coefficient = 0.633 and Mander's coefficient = 0.9 (Fig. 2). Similar levels of partial co-localization were also seen in Cos7 cells (data not shown).

RUNX1b together with CBFβ forms CBF, which further interacts with other transcription factors, co-activators or co-repressors on RUNX1 binding elements and regulates transcription of target genes. The IL2 gene promoter harbours three RUNX1 binding sites and is reported to be regulated by RUNX1b in T cells [19]. To investigate whether 3b-RUNX1b interaction leads to the recruitment of 3b on RUNX1 binding elements on the endogenous IL2 promoter, ChIP assays were performed in RUNX1b/CBFβ endogenously expressing jurkat cells that are abortively infected by SARS-CoV. Jurkat cells transfected with empty vector or HA-3b were subjected to the assay after 48 h of transfection. ChIP results from 3b transfected cells using anti-HA depicted co-immunoprecipitation of the IL2 promoter region containing RUNX1 binding site but not from mock transfected cells (Fig. 3). However, the 3′-distal IL2 promoter region, which does not contain RUNX1 binding site, was not immunoprecipitated by anti-HA in either of the samples, suggesting that 3b recruitment on the IL2 promoter region is specific to the RUNX1 binding. Control ChIP assays using anti-RUNX1 (positive control) and no antibody (negative control) were conducted simultaneously. These results clearly demonstrate an in vivo recruitment of 3b on the RUNX1 binding element on the IL2 promoter.

Viruses employ various strategies to manipulate activities of host transcription factors in their favour. To investigate the effect of SARS-CoV 3b protein on the RUNX1b transcriptional activity, reporter gene assays were performed using the mouse IL2 promoter. HEK293 cells were co-transfected with wild-type (WT) IL2 promoter luciferase plasmid, Flag-RUNX1b, Flag-CBFβ and Myc-3b in combinations mentioned. Renilla luciferase plasmid (pRL-TK) was co-transfected as an internal control. 3b alone showed no significant increase in the luciferase activity whereas 3b along with RUNX1b and CBFβ showed increase in the luciferase activity in a dose dependent manner (Fig. 4A). Transfection in jurkat cells also showed 1.5–5.0 fold increase in the luciferase activity with increasing doses of Myc-3b (Fig. 4B), suggesting that the increase in the luciferase activity in HEK293 and jurkat cells was specific to the 3b expression. To confirm whether 3b induced increase in the IL2 promoter activity was due to RUNX1b binding on the promoter, the IL2 promoter with all three RUNX1 binding sites mutated (mutant IL2-Luc) was used. Myc-3b transfected jurkat cells showed merely 0.3 fold increase in the luciferase activity with mutant IL2-Luc as against 2.0 fold with WT IL2-Luc (Fig. 4C); confirming that 3b dependent increase in the IL2 promoter activity relies on the RUNX1b transcription factor complex binding. Hence, we conclude that 3b potentiates RUNX1b transcriptional activity on the IL2 promoter.

RUNX1b transcriptional activity is regulated by various post-translational modifications like phosphorylation, acetylation, ubiquitination etc [20]. Phosphorylation of RUNX1b by Ser-Thr kinases/ERK on serine 249/266 has been reported to potentiate its transactivation potential [21]. SARS-CoV 3b has been reported to activate ERK pathway [12]. Therefore, determination of ERK dependent phosphorylation levels of RUNX1b in 3b expressing cells was of our interest. To investigate this, an in vitro kinase assay with 3b and vector transfected HEK293 cells was performed. RUNX1b immunocomplex from transfected cells served as a substrate for the kinase assay. Results depicted a 1.5 fold increase in phosphorylated RUNX1b levels in 3b transfected cells compared to control transfected cells (Fig. 5A). This suggests that 3b activated ERK may partly be responsible for the stimulated RUNX1b transcriptional activity in 3b transfected jurkat cells. To test this hypothesis, the reporter gene assay was performed in the presence of ERK inhibitor, U0126. 3b transfected jurkat cells were treated with DMSO or U0126 for 24 hrs prior to measurement of luciferase activity. Treatment of cells with U0126 led to the inhibition of luciferase activity which was otherwise observed in 3b expressing DMSO treated cells (Fig. 5B). This experiment points out that 3b mediated increase in RUNX1b transactivation potential may also be contributed by stimulated ERK activity.

Promoter of macrophage inflammatory protein (MIP-1α), a well characterised pro-inflammatory cytokine [22] have been documented to be regulated by RUNX1b through its proximal and distal RUNX1 binding elements [23]. SARS-CoV infection has been found to stimulate monocyte derived-dendritic cells to express MIP-1α [24]. In order to study the effect of 3b expression on endogenous MIP-1α mRNA levels, real-time PCR was performed in human monocytic U937 cells. Cells overexpressing RUNX1b showed 3.0 fold increase in MIP-1α mRNA levels; whereas those expressing 3b showed 5.0 fold increase as compared to vector. Interestingly, cells overexpressing 3b along with RUNX1b showed 13 fold increase in MIP-1α mRNA levels (Fig. 6), suggesting that 3b significantly enhances RUNX1b transactivation potential on the endogenous MIP-1α promoter.

3-Aminotrizole was purchased from Sigma. Luciferase assay kit (E1500) and in vitro transcription and translation kit (L4610.) were purchased from Promega. TRIzol was purchased from Invitrogen. Antibodies against RUNX1, HA and c-Myc were purchased from Santa Cruz. Anti-Flag was purchased from Roche. Alexa 488 conjugated anti-mouse and alexa 594 conjugated anti-rabbit were purchased from Molecular Probes. The SARS-CoV 3b gene (25689–26153 bp), was PCR amplified from the SARS-CoV genome (NC_004718) and cloned in pXJ40-HA vector as described earlier [56]. The 3b insert was further cloned into pHybLexA/Zeo and pCDNA3.1(-) myc/his vector. pCMVFlag Tag 2B-RUNX1b, mouse WT IL2-Luc and mutant IL2-Luc (all three RUNX1 binding sites mutated) constructs were generously provided by Dr. Shimon Sakaguchi (Institute for frontier Medical Sciences, Kyoto University, Japan), MigRI-CBFβ was gifted by Dr. Nancy A Speck (University of Pennsylvania School of medicine, Philadelphia).

Lex A based screening system (Hybrid hunter, version F), comprised of Saccharomyces cerevisiae strain L40 [MATa his3Δ200 trp1-901 leu2-3112 ade2 LYS2::(4lexAop-HIS3)URA3::(8lexAop-lacZ) GAL4], pHybLexA/Zeo and pYesTrp2 as binding domain and activation domain vectors, respectively and human lung cDNA library cloned in pYesTrp2, was purchased from Invitrogen. Screening was performed as per manufacturer's protocol. The bait plasmid was constructed by cloning 3b coding sequence in-frame with the LexA DNA binding domain in pHybLexA/Zeo. pHybLexA/Zeo-3b was co-transformed with cDNA library in L40 and co-transformants were selected for the activation of two reporter genes, HIS3 and LacZ. Strength of the interaction in selected co-transformants were assessed by their ability to grow on His⁻ Trp⁻ and Zeo⁺ YC media supplemented with 5 mM AT (3-amino-1,2,3-trizole, competitive inhibitor of HIS3) and for positivity of filter β–galctosidase activity assay. Filter-lift assay was performed as described before [57]. Plasmids were isolated from positive co-transformants and shuttled into E.Coli DH5α and sequenced. L40 co-transformed with pHybLexA/Zeo-Fos and pYesTrp2-Jun was used as the positive control and L40 co-transformed with pHybLexA/Zeo and pYesTrp2 was used as the negative control for the library screening.

HEK293, human embryonic kidney cells and Huh7, human hepatoma cells were maintained in DMEM (Dulbecco's modified Eagle's medium). Human leukemic T cells, jurkat and human monocytic cells, U937, were maintained in RPMI1640. All cell lines were maintained in media supplemented with 10% FCS (fetal calf serum, v/v) and antibiotics (penicillin and streptomycin). Transient transfections in Huh7 and HEK293 cells were carried out using fugene 6 (Roche) and lipofectamine 2000 (invitrogen) as per manufacturer's protocol and in jurkat and U937 cells were carried out by electroporation at 260 V and 975 µF in 4 mm cuvettes, using Biorad Gene Pulser.

For immunofluorescence assay, HEK 293 cells were seeded at 50% confluency on coverslips in a 12 well plate. Cells were singly or co-transfected with RUNX1b (pCMV-Tag2B Flag RUNX1b) and 3b (pXJ40 HA-3b) expression vectors and the assay was performed 24 h post-transfection as described by Korkaya et al. [58]. 3b and RUNX1b was stained using mouse anti-HA and rabbit anti-RUNX1 antibodies, respectively. Alexa 594 conjugated goat anti-rabbit and alexa 488 conjugated goat anti-mouse were used as secondary antibodies. Coverslips were mounted on to the glass slides in anti-fade (with DAPI) medium. Confocal images were collected using a 60× objective on a Nikon A-1Rconfocal microscope and analysed by imaging software, NIS-elements.

For in vitro co-immunoprecipitation, [³⁵S] radiolabelled full length RUNX1b and 3b were expressed using a coupled in vitro transcriptional/translation system as per manufacturer's protocol. 3b lysate, RUNX1b lysate and 3b-RUNX1b lysate mix were incubated with anti-RUNX1 in 500 µl lysis buffer (20 mM Tris pH 7.4, 150 mM NaCl, 0.05% NP40) at 4°C for 4–6 h. Immunocomplexes were precipitated by adding 50 µl sepharose-A (50% v/v) beads at 4°C for 1 h. Beads were washed thrice with lysis buffer and resuspended in SDS-PAGE loading buffer. Samples were run on SDS-PAGE and bands were detected by autoradiography. For in vivo co-immunoprecipitaion, Huh7 cells were transiently transfected with expression plasmids of RUNX1b (pCMV-Tag2B Flag RUNX1b) and 3b (pCDNA3.1 Myc-3b) in combinations mentioned. Total DNA amount was normalized by adding pCDNA3.1. 48 h post transfection, cells were lysed and immunoprecipitation and western blotting was performed as described earlier [59].

For ChIP, 10–15×10⁶ Jurkat cells were transfected with 10 µg vector or 3b (pXJ40 HA-3b) plasmid. 48 h post transfection, ChIP assay was performed as described earlier [60]. DNA fragments were analysed for the recruitment of RUNX1b and 3b on the RUNX1 binding site on the IL2 promoter as well as on 5′ distal region of the human IL2 gene (non-RUNX1 binding site). Sequences of the primers used for PCR: forward primer for human IL2 promoter: 5′-CTCTAGCTGACATGTAAGAAGC-3′; reverse primer for human IL2 promoter: 5′-CTACACTGAACATGTGAATAGC-3′; Forward primer for 3′ distal region of the human IL-2 gene: 5′-AAATGTTGCAGGATCCTTGC-3′; reverse primer for 3′ distal region of the human IL-2 gene: 5′-TGAGCTCTGACATGATGCTC-3′.

Luciferase assay was performed as per manufacturer's protocol (Promega). For the assay, cells were transfected with reporter plasmid (WT IL2-Luc or mut IL2-Luc), pRL-TK and indicated expression plasmids. pEGFP-N1 plasmid was transfected as a negative control in cells not expressing 3b. Drug treatment (U0126, 10 µM) was performed for 24 h prior to harvesting. All transfections were normalized for the amount of total DNA by adding pCDNA3.1. Luciferase activity was measured 48 h post transfection and normalized against renilla luciferase activity. Relative luciferase activity was expressed as mean±standard deviation (SD) of three independent experiments. Statistical significance was calculated using student's t test. p values are given in the figure legends. Error bars represent the SD.

ERK activity in 3b transfected cells was assayed using immunoprecipitated RUNX1b as the substrate. HEK293 cells were transfected with expression plasmids of vector (control), 3b, and RUNX1b. 48 h after transfection, cells were lysed in lysis buffer (20 mM Tris-HCl pH 7.5, 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, 1% Triton, 2.5 mM sodium pyrophosphate, 1 mM β-glycerophosphate, 1 mM sodium orthovandate) and equal amounts of vector and 3b lysates were subjected to immunoprecipitaion using anti-ERK antibody and RUNX1b lysate, using anti-RUNX1 antibody. Sepharose A beads were washed thrice with the lysis buffer and twice with the kinase buffer (20 mM Tris–Cl pH 7.5, 10 mM MgCl₂, 1 mM DTT, 5 mM Sodium orthovandate). For the assay, ERK beads were incubated with RUNX1b beads in 20 mM Tris–Cl (pH 7.5), 10 mM MgCl₂, 1 mM DTT, 2 mM β-glycerophosphate and 50 mM ATP, 5 mCi [γ-³²P]-ATP at 37°C for 30 min. Reaction mixtures were run on SDS–PAGE, and analyzed by autoradiography.

RNA was isolated from transfected U937 cells using TRIzol, as per manufacturer's protocol. 5 µg RNA was reverse transcribed using oligo-dT primer and MuLV reverse transcriptase as per manufacturer's protocol (Promega). Primers used for PCR were: MIP-1α forward: 5′ ACTTGCTGCTGACACGCCGA 3′ and reverse: 5′ CACAGACCTGCCGGCTT CGC 3′; Actin forward: 5′ TGACGGGGTCACCCACACTGTGCCCATCTA3′ and reverse: 5′ CTAGAAGCATTTGCGGTGGACGATGGAGGG3′. Data is represented as bar graph and is mean±SD of three independent observations.